THU0131 Mirna Profiling Using HTG-Edgeseq Platform Predicts Response To anti-TNF Alpha Therapy in Rheumatoid Arthritis

Background Millions of patients suffering Rheumatoid Arthritis (RA) are treated with TNF-α inhibiting agents, however, response rate is low (30 to 40%) and no tool exists to predict the treatment response. Objectives Using the HTG-Edgeseq platform, an innovative combination of nuclease protection assay and next generation sequencing, we identified sets of miRNAs that discriminate responders from non-responders to anti TNF-α therapy. Methods Sixty-seven patients diagnosed with RA, eligible for treatment with 1st line anti-TNFα and for whom DMARD therapy had failed were enrolled in the study. Twelve to 14 weeks after anti-TNFα therapy, patients were categorized as responders or non-responders based on DAS28 index. Patients9 miRNA profile was established from 15μl of plasma using HTG-Edgeseq Whole Transcriptome Assay (WTA) miRNA panel (2256 miRNA). Results were normalized based on the median of the sample and Random forest was used as the classification model. For 8 patients, miRNA were also analyzed with the qPCR Exiqon miRNA panel V4 to determine HTG-Edgeseq accuracy. Method reproductibility was assessed by analyzing 4 times an independent sample on different sites, with different instruments/days/operators. Results Results obtained from both HTG-Edgeseq and qPCR methods showed an overall correlation of 0.63 for the 341 miRNA common between those 2 kits, and correlations factors between the 4 independent experiments ranged from 0.993 to 0.999. Statistical analysis of patients9 miRNA profile identified 2 panels of 6 and 52 miRNAs with significant predictive power to discriminate responders from non-responders (sensitivity was 0.898 and 0.918, and AUC 0.773 and 0.824 respectively). Conclusions In conclusion, miRNA profiling in RA patients using HTG-EdgeSeq allowed us to build 2 predictive models for response to anti-TNF-α drugs. Moreover, we showed that HTG-Edgeseq platform offers accurate and sensitive miRNA expression measurement. Its low sample input requirement and compatibility with all biological material makes it an invaluable tool for biomarkers discovery. Disclosure of Interest None declared